i-R2D247 (~deed@tor-irc.dnsbl.oftc.net) left irc: Remote host closed the connection
i-R2D247 (~deed@tor-irc.dnsbl.oftc.net) joined #fai.
meandtheshell (~markus@ joined #fai.
alleeHol (~ach@dialin-212-144-130-005.pools.arcor-ip.net) joined #fai.
07:23 alleeHolmorgen
sep (~sep@ joined #fai.
Mrfai (~lange@kueppers.informatik.uni-koeln.de) joined #fai.
alleeHol (~ach@dialin-212-144-130-005.pools.arcor-ip.net) left irc: Read error: No route to host
alleeHol (~ach@dialin-212-144-130-252.pools.arcor-ip.net) joined #fai.
Nick change: alleeHol -> allee
sep (~sep@ left irc: Ping timeout: 480 seconds
sep (~sep@ joined #fai.
10:47 mikap"Having stared and poked at FAIs new partition framework makes me consider partman as an alternative." -- http://schuldei.blogspot.com/2007/10/dear-lazyweb-au-pair-wanted.html - hm, what does he mean? :)
10:49 stockholmmikap: the control flow is not really designed to be capable of handlinge raid and lvm
10:50 stockholmmikap: preserve in this context makes only sense for partitions currently
10:51 mikapstockholm: ah ok
10:51 stockholmi tried to hack it into removing first and building new stuff then, but that is hard becuase of the control flow
10:52 stockholmor the lack thereof
10:53 stockholmall this is done beautifully in partman. it lacks an interface that fai can use
10:54 stockholmand it needs to be made non-interactive
11:04 Mrfaiand partman must be independent of d-i.
11:04 Mrfaistockholm: but I do not understand what you mean by "control flow"?
11:05 stockholmMrfai: it must have an interface that we can use
11:05 stockholmMrfai: with controlflow i mean several things
11:06 stockholmthe most obvious is that there is no seperation for tearing down and building up lvm, raid and partitions seperatily
11:06 stockholmit handles lvm, raid and partitoins in one go
11:07 Mrfaiis the problem that the partition tool does all in one go, or the the description is in one file?
11:07 Mrfais/the the/that the/
11:08 stockholmthat means it first removes partitions, creates new, THEN (after having broken its basis) removes raids, creates new ones, THEN (after hanving broken its basis) attempts to create, resize and/or delete lvm volumes
11:08 stockholmdescription in one file is no problem
11:08 stockholmthe way the actual work takes place is bogous
11:09 stockholm(the -commands file)
11:09 stockholmthere is a delicate interaction between partitions, raids and lvms on top
11:10 Mrfaiok. Would it help to add options, so it only does the raid or lvm thing or only the partition thing, or multiple things in one go?
11:10 stockholmit needs to get smarter to reorder things.
11:11 stockholmi tried to do that in a quick approach
11:11 stockholmbut that failed because the teardown and buildup as it is now relies on the same data structures and changes them of course as it goes
11:12 stockholmmy apprach was to have a "down" and "up" mode and permit only teardown commands in the "down" mode etc
11:12 stockholmbut that fails obviously
11:13 stockholm"preserve" needs to be rethought, too.
11:13 stockholmas it makes most sense to preserve the *data*, not the partitions
11:14 stockholmit is possible to preserve all the data in lvms but change the partition layout
MT (~MT@M247P017.adsl.highway.telekom.at) joined #fai.
11:14 stockholmMT: hi
11:14 MThi!
11:15 stockholmMT: http://paste.debian.net/40220
11:15 MrfaiMT: http://paste.debian.net/40219, this is what happens in the last minutes
11:15 stockholmduh
11:15 Mrfai:-)
11:15 stockholm:-)
11:16 Mrfaistockholm: do we need to split preserve in preserve-data and preserve-partition?
11:17 stockholmMrfai: we need to make it different things if lvm is involved at least
11:17 stockholmmake it do
allee (~ach@dialin-212-144-130-252.pools.arcor-ip.net) left irc: Read error: No route to host
11:18 stockholmMT: i added a bit of what i have been thinking to the wiki
11:18 stockholmand filed a RFP for LINUX::Lvm
11:19 stockholmMT: it would be smart to re-implement the relevant partman bash stuff in perl for our purpose
11:20 stockholmMT: or write a parser and logger to partman in bash
11:20 MTwhich is the "relevant" partman part?
11:20 MTI guess partman does not have any sort of preserve/resize means, does it?
11:20 stockholmthe logic for preserving, resizing, removing and creating of lvm, raid and plain partitions
11:21 stockholmit can resize everything that moves
11:21 stockholmor is on disk :-)
11:21 stockholmeven advancd stuff with lvm and raid involved
11:21 stockholmor windows partitions i think
11:21 stockholmit can not sanely handle multiple disks yet
11:22 stockholmit only cares about the first disk at this point
11:22 stockholmMT: i think the parser in shdd2 is very good and exactly what we need
11:23 stockholmi have not seen much else besides -commands
11:24 MTI think that changing the sequence of commands shouldn't be all that hard
11:24 MTcurrently there is a single list FAI::commands
11:25 stockholmgood
11:25 MThmm, resolving dependencies between lvm and RAID might be an issue
11:25 MTbut anyway
11:25 MTif we had two lists instead
11:26 MTthe tear-down commands and the build-up
11:26 stockholmthat would be a start
11:27 stockholmhandling the tear down first, handling errors, starting build up based on that
MT (~MT@M247P017.adsl.highway.telekom.at) left irc: Ping timeout: 480 seconds
MT (~MT@M199P014.adsl.highway.telekom.at) joined #fai.
12:42 oz_moin
MT (~MT@M199P014.adsl.highway.telekom.at) left irc: Ping timeout: 480 seconds
MT (~MT@dove.informatik.tu-muenchen.de) joined #fai.
meandtheshell (~markus@ left irc: Quit: Leaving.
allee (~ach@dialin-212-144-132-050.pools.arcor-ip.net) joined #fai.
MT (~MT@dove.informatik.tu-muenchen.de) left irc: Ping timeout: 480 seconds
kugg (~jerkeby@ joined #fai.
meandtheshell (~markus@ joined #fai.
allee (~ach@dialin-212-144-132-050.pools.arcor-ip.net) left irc: Read error: No route to host
allee (~ach@dialin-212-144-131-045.pools.arcor-ip.net) joined #fai.
MT (~MT@dove.informatik.tu-muenchen.de) joined #fai.
MT (~MT@dove.informatik.tu-muenchen.de) left irc: Ping timeout: 480 seconds
MT (~MT@dove.informatik.tu-muenchen.de) joined #fai.
MT (~MT@dove.informatik.tu-muenchen.de) left irc: Ping timeout: 480 seconds
MT (~MT@dove.informatik.tu-muenchen.de) joined #fai.
meandtheshell (~markus@ left irc: Quit: Leaving.
MT (~MT@dove.informatik.tu-muenchen.de) left irc: Ping timeout: 480 seconds
allee (~ach@dialin-212-144-131-045.pools.arcor-ip.net) left irc: Ping timeout: 480 seconds
Lin (~igor@ joined #fai.
cesar (~cek@ joined #fai.
21:58 stockholmhi cesar
21:58 stockholmhi from sweden
21:59 cesarhi from Extremadura
22:01 stockholmcesar: how is the weather there?
MT (~MT@dove.informatik.tu-muenchen.de) joined #fai.
22:02 stockholmMT: hi
22:02 MThi!
22:02 stockholmMT: ich hoffe ich habe dich nicht frustriert mit meiner meinung, ich werde wohl weiter daran arbeiten
22:03 stockholmMT: ich bin mir nur nicht sicher welcher ansatz weniger arbeit ist
22:04 MTstockholm, ich bin mit allem einverstanden, was zu einem guten Paritionierungstool führt
22:04 MTaber jetzt werde ich erstmal schlafen :-)
22:04 stockholmMT: ja, gute nacht
22:05 cesarstockholm, the weather is fresh and nice :)
22:06 cesarich kann auch Deutsch schreiben, aber nicht zu viel...
MT (~MT@dove.informatik.tu-muenchen.de) left irc: Ping timeout: 480 seconds
22:17 stockholmcesar: i can write no spanish at all
22:17 stockholmcesar: i am told you host debian developer gatherings in extremadura :-)
22:18 cesarwho told you that? ;)
22:19 stockholmcesar: let me check
22:19 seph01ger,
Lin (~igor@ left irc: Quit: Ex-Chat
22:20 h01gerstockholm,
22:20 h01gersep,
22:20 stockholmcesar: it was "César Gómez Martín" <cesar.gomez@gmail.com>
22:21 cesarlol ok, so it was me then :)
22:21 stockholmi think so
22:22 stockholmh01ger: ?
22:23 h01gerstockholm, ?
22:26 h01gersep
22:33 lazyb0yhey cesar :)[6~[6~[6~
22:33 lazyb0y34433
22:34 cesarwhat? :D
22:38 lazyb0yaehm - i didn't mean that - network problems, hitting some random keys
22:39 kugghehe
22:52 h01gerlazyb0y, Packages.gz  files/packages/Packages.gz
22:53 h01gers/lazyb0y/eartoast/
22:56 h01germootbot_chan.tcl  files/home/meetbot/scripts/mootbot_chan.tcl
22:56 h01germootbot_config.tcl  files/home/meetbot/scripts/mootbot_config.tcl
Action: oz_ back from finishing his third or forth FAI installation
sep (~sep@ left irc: Quit: Leaving
cesar (~cek@ left irc: Quit: Abandonando
--- Sun Oct 21 2007

Generated by irclog2html.pl 2.1_h01ger1 by Jeff Waugh - find it at freshmeat.net! Customized by Holger 'h01ger' Levsen